When a public organization decides its communication team needs more analytical capacity, the usual reflex is to hire "someone who does data." That someone tends to end up being a strong technical profile — statistics, data engineering or data science — who produces elaborate dashboards the communication team doesn't know how to use, and models nobody asked for.

The problem isn't the professional who was hired. It's that a profile was hired to solve a problem that profile doesn't solve. A classic classification from the field identifies nine types of data analysts. Let's look at which ones fit public communication and, above all, which one doesn't.

The nine profiles, summarized

Drawn from the data science field's own literature, the classification distinguishes:

  1. Applied statistician: specializes in statistical models, experimental design, clustering, predictive modeling.
  2. Applied mathematician: operations research, process optimization, deep quantitative analysis.
  3. Data engineer: Hadoop, databases, APIs, data pipelines, infrastructure ("data plumbing").
  4. Machine learning scientist: machine learning, algorithms, computational complexity.
  5. Business analyst: ROI, decision science, applied financial analysis.
  6. Software developer: production code, software engineering for analytical systems.
  7. Visualization specialist: dashboards, visual exploration, data storytelling.
  8. Spatial / GIS analyst: spatial data infrastructure, graphs, geolocation.
  9. Generalist: a pragmatic combination of several of the above.

Which one gets hired most often, and why it's the wrong one

Because of their reputation, media visibility and the volume of training programs built around them, the most commonly hired profiles in recent years have been the machine learning scientist and the data engineer. Both are excellent professionals. Both are, in most public communication teams, the wrong choice.

The reason: the dominant problem in institutional communication isn't building a new predictive model or standing up a complex data infrastructure. It's turning existing data into decisions that non-technical people can actually make. That calls for a different profile.

Hiring a data scientist to solve a non-technical decision-making problem is like hiring an architect when what you need is to decorate the living room. The sophistication is there — it just doesn't solve the problem.

The one that does fit: the business analyst with a visual sense

The profile that best fits a public organization's communication team combines two of the nine types: business analyst and visualization specialist. The operational reasons:

Why this mixed profile works

  • Speaks the language of decision-making. Their training is oriented toward supporting decisions, not building models. This shortens the distance between data and action.
  • Can explain data without jargon. Visualization isn't decoration: it's the interface through which a non-technical manager will consume the analysis.
  • Spots what matters without getting locked into tools. No emotional attachment to a particular technology stack; uses whatever works.
  • Knows the limits of the data. Distinguishes operational metrics from decisive ones, avoiding dashboards packed with numbers that inform nothing.

When it does make sense to hire an advanced technical profile

There are three situations where a technical profile — data engineer, machine learning, applied statistician — is the right choice:

  1. When there's a data infrastructure to build from scratch. If the organization doesn't yet have the basic pipelines in place, hiring a visualization specialist is premature.
  2. When there's a clear, recurring predictive problem. For example, anticipating reputational crises with some lead time, or predicting the behavior of a segment. This justifies an ML profile.
  3. When the volume and heterogeneity of sources demands specialized engineering. Large-scale multilingual digital listening, integration of many APIs, real-time sources.

Outside these three cases, hiring an advanced technical profile to support communication is over-engineering that doesn't pay off.

The opposite error: hiring a visualizer with no analytical judgment

There's also the opposite mistake: hiring someone who builds pretty dashboards but doesn't understand which metrics matter or why. The result is aesthetics without substance — very presentable panels that fall apart under the first serious committee review, because they show numbers that decide nothing.

The rule: visualization should be the last step, not the first. First you decide what needs deciding, then which data informs it, then how it's visualized for the specific audience that will consume it.

What we ask an analytical profile before bringing them on board

Three questions we use when bringing analytical profiles into communication teams:

  1. "Tell me about a real decision that changed because of an analysis you did." If they can't tell it in detail, they haven't worked in environments where the data actually mattered.
  2. "Show me a dashboard you built and tell me what you'd remove now." Someone who doesn't critique their own work hasn't learned enough from it.
  3. "How would you explain this metric to someone non-technical?" If they can't, they probably don't use it to inform decisions either.

If you're about to strengthen your communication team's analytical capacity and want to make sure you hire the right profile for your case, that's exactly the kind of outside judgment we can bring.

Let's talk

No commitment. Reply within 48h.

Conclusion

Analytical hiring in public communication teams is a field where the obvious profile is rarely the right one. Technical sophistication doesn't solve a problem that is, at bottom, about translating data into decisions.

The mixed profile — business analyst plus visualization — is, in most institutional cases, the one that fits best. The advanced technical profile has its place, but only when there's a real technical problem that justifies hiring it. Getting this choice right can save one or two years of misdirected investment.